Policy Intervention Is Keeping The Bull Market Alive | Weekly Roundup
The Forward Guidance roundup panel discusses how Trump's geopolitical maneuvering (particularly around Iran) appears timed to stabilize markets ahead of the SpaceX IPO, while analyzing derivatives positioning, the underperformance of Mag7 stocks, and the broader implications of centralized asset management and AI policy.
Summary
The episode opens with casual discussion before diving into market analysis centered on a key observation: Trump's announcement of an Iran deal appeared strategically timed to prevent market destabilization ahead of the SpaceX IPO. The panel notes this follows a recognizable pattern from late March, where geopolitical threats are escalated and then reversed at precise moments when currency and equity markets are most vulnerable — particularly when USD/JPY approaches 160 and volatility spikes. The panel characterizes this as 'centralized asset management,' arguing that the concentration of capital in large systematic and quant funds makes markets highly susceptible to these kinds of policy interventions.
On monetary policy, the group argues that hawkish rate hike expectations are overdone. They point to falling two-year breakevens, negative real wage growth, and the absence of a wage-price spiral as evidence that inflation is primarily supply-shock driven rather than demand-driven — and thus unlikely to require Fed tightening. They argue 'max asymmetric hawkishness' has been reached and that SOFR futures represent an asymmetric long opportunity, since either an AI-fueled growth scenario requires cuts, or demand destruction from energy prices leads to the same outcome. Kevin Warsh's expected hawkishness at the FOMC is seen as overstated given his stated preference for core inflation measures.
A significant portion of the discussion focuses on the Mag7 underperformance relative to equal-weight indices, attributing it to massive equity issuance (Google ~$80B, Oracle ~$40B, SpaceX ~$75B), increased CapEx spending with uncertain returns, and the unwinding of the dispersion trade as implied correlations collapsed and then snapped back. The panel argues the smarter AI trade is to buy what hyperscalers are purchasing (memory, power infrastructure) rather than the hyperscalers themselves, as they are net issuers of equity funding those purchases.
The conversation broadens into concerns about AI centralization, specifically Anthropic's 'Fable' model restricting access to its more powerful 'Mythos' model based on approval criteria. The panel expresses concern that gatekeeping frontier AI models stifles entrepreneurial opportunity and mirrors historical patterns of institutional capture. They also discuss Bitcoin miners being approached by AI companies needing verified power capacity, the political headwinds building against unlimited AI infrastructure spending ahead of midterms, and the broader K-shaped economy dynamic where policy continues to favor asset holders while the middle class falls further behind.
Key Insights
- The panel argues Trump's Iran deal announcement was deliberately timed to prevent market destabilization ahead of the SpaceX IPO, following the same playbook used in late March — creating geopolitical fear and then reversing it precisely when USD/JPY hits ~160 and volatility spikes.
- The panel contends that 'max asymmetric hawkishness' has already been reached in rate pricing, arguing that two hikes priced into the curve are unjustified given negative real wage growth, falling two-year breakevens, and the supply-shock nature of current inflation.
- One speaker argues SOFR futures represent the most asymmetric long trade available because the outcome is binary — either the Fed cuts to support AI-driven growth, or demand destruction forces cuts anyway — making the downside of being long SOFR very limited.
- The panel identifies the unwinding of the dispersion trade as a key driver of recent volatility: implied correlation had fallen to ~6, single-stock vol was exploding while index vol was flat, and when the trade reversed, retail call buyers were 'absolutely wiped out' as index vol spiked.
- One speaker argues the smarter AI infrastructure trade is to buy what hyperscalers are purchasing (e.g., memory, which is up 3x) rather than the hyperscalers themselves, since companies like Google are issuing tens of billions in equity to fund those purchases, creating structural headwinds for their own share prices.
- The panel describes Anthropic's new 'Fable' model — a restricted version of its 'Mythos' model — as a form of regulatory capture and innovation gatekeeping, noting that users on the restricted tier are being blocked even from basic biology questions, while approved users get unrestricted access to frontier capabilities.
- One speaker observes that Bitcoin miners with verified power capacity are being aggressively approached by AI companies because many other data center operators claiming large power reserves don't actually have grid-approved capacity, making miners with operational power uniquely valuable.
- The panel argues that the concentration of capital in large systematic and quant funds has effectively created a centrally managed market, where government actors can time announcements to force CTA re-leveraging and short squeeze dynamics, and that this structure cannot change until the asset management industry decentralizes.
Topics
Full transcript available for MurmurCast members
Sign Up to Access